Noise image edge detection by synthetic weighted multi-scale morphology

Noise image edge detection is an important image processing work. Differential edge detection operators can't filter the noise effectively, and common multi-scale edge detection methods blur the edge and can't detect the details of the edge. This paper represented a novel multi-scale morphological edge detection algorithm. To utilize the noiseproof feature of larger scale structuring element to restrain the noise, and the allocation feature of smaller scale structuring element to trace and detect the edge details, synthetic weighted method was proposed to compromise the merits of different scale structuring elements. The experimental results showed that the novel proposed algorithm was more effective for noise image edge detection than differential edge detection operators and common multi-scale morphology edge detection methods.